Update semconv schema to define stability as optional for attribute groups.

[jerbly]

@lmolkova wrote:

... the attribute_group stability does not make sense - https://github.com/open-telemetry/semantic-conventions/blob/main/docs/general/group-stability.md#group-stability. It'd be really awesome if we changed yaml schema and tooling behind it to reflect this

This change updates the semconv schema and the syntax doc to correctly define stability as optional for attribute groups.

[codecov]

Codecov Report

All modified and coverable lines are covered by tests ✅

Project coverage is 72.5%. Comparing base (cab3125) to head (b65bfc2).

Additional details and impacted files

@@           Coverage Diff           @@
##            main    #467     +/-   ##
=======================================
- Coverage   72.5%   72.5%   -0.1%     
=======================================
  Files         49      49             
  Lines       3635    3635             
=======================================
- Hits        2639    2638      -1     
- Misses       996     997      +1     

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

---

🚨 Try these New Features:

• Flaky Tests Detection - Detect and resolve failed and flaky tests

[jerbly]

I noticed that AttributeFullSpec had stability as required but without the property!

[lquerel]

LGTM. Thanks